Post-processing

Key takeaways

• Cull before detailed editing
• Compare similar frames side by side
• Use one consistent flag, rating, and color-label convention
• Check critical sharpness only among final competitors, not every frame in the first pass
• Select for story and output requirements, including horizontal or vertical use
• Keep alternates separate instead of filling the main edit with near-duplicates

Practical example

From twelve similar portraits, remove blinks first, then compare expression and gesture, and retain one frame with the clearest moment. If a horizontal hero and vertical social crop serve different deliverables, label their roles; five tiny variations of one smile do not.

Key takeaways

• Set profile, white balance, and overall exposure before building a strong look
• Crop to clarify composition while preserving required output resolution
• Straighten against a line that should truly be level, not a perspective edge
• Protect meaningful highlights and shadows instead of forcing every tone into range
• Compare before/after at similar perceived brightness
• Use virtual copies or non-destructive versions to explore alternatives
• Lens correction uses the matching profile to address optical distortion and vignetting; inspect chromatic fringes at high-contrast edges. It differs from perspective correction: converging verticals mainly arise from camera orientation and need a separate transform/keystone adjustment.
• For perspective correction, choose reliable reference lines, transform only as much as needed, and crop empty edges. Leave extra framing at capture because the transform discards edge pixels. Do not force every line parallel when an upward view is intentional.

Practical example

A portrait in open shade may need a warmer white balance, +0.4 EV, and reduced highlights. A crop that removes a bright edge object keeps attention on the face. After warming, check that skin has not become orange and that the remaining file is large enough for delivery.

Key takeaways

• Set white and black points while protecting important detail
• A gentle S-curve raises contrast; a lifted black point lowers the sense of depth
• Strong curves can change perceived saturation, so reassess color afterward
• HSL should support subject color and palette, not create isolated false color
• Large luminance changes can produce halos, noise, and broken gradients
• Review at thumbnail, Fit, and 100% because each reveals different problems

Practical example

Use a subtle S-curve for shape, then reduce green saturation if foliage competes with skin. If the subject's clothing is also green, use a background mask or a narrower range instead of changing all green pixels. Temporarily view in monochrome to verify tonal hierarchy.

Key takeaways

• Feather edges enough to avoid obvious transitions
• Range masks can prevent a gradient from affecting unrelated tones or colors
• Use subtle dodge and burn to guide attention rather than remake the scene
• Several clearly named simple masks are easier to control than one complex mask
• Check for halos around hair, mountains, and high-contrast edges
• Toggle each mask and the entire edit to identify cumulative overprocessing
• Dodge & Burn selectively lightens/darkens to guide attention or smooth small tonal differences. Use a soft low-flow brush, build gradually, and return to the whole image. Preserve original light direction and texture rather than repainting form with strong Exposure changes.
• Skin retouching begins with an agreed scope. Heal/Clone temporary blemishes nondestructively, then correct slight unevenness in brightness. Retain pores and identifying features; do not assume permission to remove moles or wrinkles or alter body shape.
• For object removal, Heal blends sampled texture into its surroundings, Clone copies an area, and Remove/Generative Remove can synthesize content. Inspect edges, shadows, reflections, and repeated textures at 100% and full view. In documentary work, removing objects can change meaning and must follow the project’s publication standards.

Practical example

Raise the face about +0.2 EV, lower the sky with a gradient, and darken a bright distraction slightly. If an AI subject mask misses hair, refine it with a brush or intersection rather than widening feather until the whole background changes. For an agreed creative portrait with a temporary cheek blemish and a plastic bag at the edge, retain the original, heal from a nearby appropriate area, use subtle dodge & burn, and remove the bag on the edited version. For a documentary frame where the bag is information, keep it or choose another frame rather than silently erasing it. Inspect hair edges, the bag’s shadow, and wall texture for artifacts.

Key takeaways

• Correct exposure before judging the required noise reduction
• Reduce color noise before making strong luminance-noise decisions
• Mask sharpening away from skin, sky, and smooth backgrounds
• Small radius suits fine detail; large radius more easily creates halos
• Inspect artifacts at 100% but decide at actual output size
• Web and print exports need separate resizing and output sharpening

Practical example

For an ISO 6400 portrait, reduce color noise, add only enough luminance reduction to control distraction, and mask sharpening into eyes and hair. Compare a version that retains grain and hair strands with a cleaner version where hair begins to clump. At web size, retained texture often looks more natural.

Key takeaways

• Make a neutral correction before a creative grade
• Describe a reference through tone, palette, saturation, and color relationships
• Keep skin and important recognition colors believable
• Synchronize visual intention, not every numeric slider
• Test a style across multiple scenes and lighting conditions
• Use a calibrated, stable viewing environment when color accuracy matters
• For skin tone correction, establish WB/Tint from a reference first, use a skin mask for local casts, and adjust Hue/Saturation/Luminance gradually while comparing face, neck, and hands. No single HSL value or brightness target is correct for every complexion.
• Colour grading deliberately colors shadows, midtones, and highlights after neutral correction. Use a reference to choose cool shadows or warm highlights, then reduce the effect if skin or product colors drift.
• A preset stores processing settings; a LUT maps color values and usually expects a specific input color space/gamma. Neither replaces WB, exposure, or the input profile. A Log-input LUT should not be applied indiscriminately to already rendered sRGB JPEG/RAW output.
• For batch editing, group frames by lighting, develop a reference, and synchronize only shared settings. Exclude framing-specific crop, heal, and positional masks by default, recheck adaptive masks, and refine individual exposure/WB.
• Colour consistency aligns visual feeling and identity colors rather than forcing equal histograms or Kelvin values. Inspect a contact sheet, establish a reference per lighting group, and correct outliers before exporting the album.
• Develop signature colour across projects: write three principles for tone, palette, and protected colors, version the preset, and test sun/shade/artificial light. Record where it fails. A personal style grows from intentional repeated choices, not simply one LUT.

Practical example

A “warm, muted, soft” direction may use slightly warm highlights, lower overall saturation, restrained greens, believable orange skin, and gentle contrast. A sunlit file and a shade file may need different WB and HSL values to reach that same destination. For ten sunlit and ten LED-lit images, choose one reference per group, correct green skin casts, then add gentle warm grading. Sync curves and palette within each group, excluding crop/heal. Compare face, neck, hands, and the contact sheet; correct orange drift in an individual frame rather than changing the whole preset. Save “warm-soft-v01” with suitable lighting and known failure cases.

Key takeaways

• Select software according to workflow bottlenecks rather than popularity
• Learn editing principles before relying on presets and plugins
• Keep adjustments non-destructive where possible
• Back up the catalog/database separately from RAW files and sidecars
• Layered PSD/TIFF masters can be large, so round-trip only images that need them
• Evaluate AI and plugins for artifacts, privacy, and reproducibility
• A RAW workflow backs up originals and catalog/sidecars, culls, sets a profile, corrects lens/WB/exposure, refines tone/color, applies masks and detail work, then exports copies for the destination. Retain history or virtual copies so style changes do not overwrite originals.

Practical example

For 500 images, cull and batch-correct in Lightroom or Capture One. Send only five hero images needing complex cleanup, compositing, or retouching into Photoshop as layered masters, then stack them beside their RAW files in the catalog.